Skip to main content
. 2015 Aug 11;351:h3868. doi: 10.1136/bmj.h3868

Estimates of regression coefficients, calculated by standard regression and penalised methods

Predictors Descriptive statistics† Regression coefficient estimates*
Standard regression Ridge regression Lasso regression
Intercept −7.80 −5.97 (23) −6.65 (15)
Sex (female) 1337 (43) −0.24 −0.14 (41) −0.16 (34)
Age (years) 54.1 (10.8) −0.052 −0.047 (11) −0.050 (4)
Body surface area (m2) 1.6 (0.3) 1.98 1.52 (24) 1.75 (12)
Aortic size 23, 27, 29, 31 mm 692 (22) 1.43 0.36 (75) 0.61 (68)
Mitral size 23-27 mm 369 (12) 1.30 0.22 (84) 0.43 (67)
Mitral size 29 mm 611(20) 1.95 0.80 (59) 1.13 (42)
Mitral size 31 mm 656 (21) 2.62 1.38 (47) 1.77 (33)
Mitral size 33 mm 104 (3) 2.58 1.41 (45) 1.73 (33)
Fracture in batch (yes) 1108 (35) 0.59 0.69 (−17) 0.64 (−9)
Date of manufacture (after 1981) 2363 (76) 1.38 1.02 (26) 1.22 (12)

*For ridge and lasso methods, numbers in brackets are percentages and represent the shrinkage compared with the standard regression estimates.

†For descriptive statistics, data are mean (standard deviation) for continuous predictors (age and body surface area), and number (percentages) for binary predictors.